Noise Floor Reduction in Frequency Delta-Sigma Modulation Microphone Sensors

Area of Science:

• Electronics
• Sensor Technology
• Acoustics

Background:

• Classical delta-sigma modulators have limitations for sensor integration.
• Frequency delta-sigma modulators (FDSMs) offer intrinsic analog-to-digital converter (ADC) integration with sensors.
• Previous FDSM applications suffered from significant signal-to-noise ratio (SNR) degradation due to noise floor.

Purpose of the Study:

• Investigate the origin of the noise floor in FDSM microphone sensors.
• Identify methods to improve the noise floor and SNR of FDSM sensors.
• Demonstrate the potential of FDSM technology for various physical sensors.

Main Methods:

• Analyzed the noise floor in FDSM microphone sensors.
• Improved the quality factor (Q-factor) of the cavity resonator.
• Optimized the design of the variable frequency oscillator circuit.

Main Results:

• Reduced the noise floor by approximately 40 dB through phase noise improvement.
• Achieved an SNR of 57 dB for 114 dBSPL sound input with a 96 kHz bandwidth.
• Obtained a dynamic range of 87 dB for maximum 140 dBSPL, with potential for 120 dB.

Conclusions:

• Improving oscillator phase noise is critical for reducing FDSM noise floor.
• Enhanced FDSM microphone sensors demonstrate significant performance improvements.
• FDSM technology shows strong potential for a wide range of physical sensor applications.